THE PROGNOSTIC VALUE OF NEURON-SPECIFIC ENOLASE IN HEAD
Emine Meric, MD,* Abdulkadir Gunduz, MD,† Suleyman Turedi, MD,† Ertugrul Cakir, MD,‡
and Mustafa Yandi, MD§
*Department of Emergency Medicine, Ordu General Hospital, Ordu, Turkey, †Department of Emergency Medicine, Karadeniz Technical
University Faculty of Medicine, Trabzon, Turkey, ‡Department of Neurosurgery and §Department of General Surgery, Karadeniz
Technical University Faculty of Medicine, Trabzon, Turkey
Reprint Address: Suleyman Turedi, MD, Karadeniz Teknik Üniversitesi Tıp Fakültesi Farabi Hastanesi, Acil Tıp Anabilim Dalı,
Trabzon 61080, Turkey
e Abstract—In recent years, in addition to neurological
examination and neuroradiologic examinations, attempts
have been made to assess the severity of post-traumatic
brain injury and to obtain an early idea of patient prognosis
using biochemical markers with a high degree of brain
tissue specificity. One such enzyme is neuron-specific eno-
lase (NSE). This study investigates the correlation between
serum NSE levels, Glasgow Coma Score, and prognosis
measured by Glasgow Outcome Scores in head trauma
patients. This was a prospective study conducted with 80
trauma patients presenting to the Emergency Department.
Patients were divided into four groups. The first group
consisted of patients with general body trauma, but no head
trauma. The second group had minor head trauma. The
third group had moderate head trauma, and the fourth
group had severe head trauma. The relationship between
subjects’ admission NSE levels and admission and dis-
charge Glasgow Coma Scores (GCS) and Glasgow Outcome
Scores (GOS) 1 month later was examined. A receiver
operating characteristic (ROC) analysis was performed us-
ing a serum NSE cutoff level of 20.52 ng/mL and a GOS of
3 or less as the definition of poor neurologic outcome. There
was a significant difference in the NSE levels between group
1 (general trauma) and group 3 (moderate head trauma).
There was also a statistically significant difference in NSE
levels between group 1 (general trauma) and group 4 (se-
vere head trauma) (p < 0.05). There was a statistically
significant inverse relationship between NSE levels and
GOS as determined within groups 3 (moderate) and
4 (severe head trauma) (p < 0.05). When NSE levels were
compared with admission GCS, it was found that GCS fell
as NSE levels rose. There was no significant correlation
between NSE and GCS within groups 3 (moderate) or
4 (severe). There was a statistically significant correlation
within group 2 (mild) (p < 0.05). By ROC analysis, serum
NSE was 87% sensitive and 82.1% specific in predicting
poor neurologic outcome in the study patients. The area
under the curve was 0.931. This study shows that initial
serum NSE levels in moderate and severe head trauma
patients correlate inversely with GOS 1 month later, but
only within the moderate and severe head trauma groups.
However, serum NSE was 87% sensitive and 82.1% specific
in predicting poor neurologic outcome in all of the study
patients. This derived cutoff value now needs to be prospec-
tively validated.© 2010 Elsevier Inc.
e Keywords—head trauma; neuron-specific enolase; prog-
The Glasgow Coma Score (GCS) is the usual method of
neurological evaluation in head trauma patients in the
Emergency Department (ED). However, initial com-
puted tomography (CT) scan findings and GCS values do
not always correlate with final outcomes (1,2). Research
RECEIVED: 14 January 2007; FINAL SUBMISSION RECEIVED: 25 July 2007;
ACCEPTED: 2 November 2007
The Journal of Emergency Medicine, Vol. 38, No. 3, pp. 297–301, 2010
Copyright © 2010 Elsevier Inc.
Printed in the USA. All rights reserved
0736-4679/08 $–see front matter
has shown that there is no relation between early CT
findings and subsequent neurological examination find-
ings (2). Moreover, patients with serious head trauma
usually require early intubation, sedation, and paralysis,
which at times can make a reliable neurologic examination
impossible for neurosurgeons and emergency physicians.
Due to the inability of early physical examination and
radiological studies to predict patients’ prognoses, there
have been recent efforts to find biochemical markers that
prognosis. Two markers with a high degree of brain tissue
specificity are S-100B, neuron-specific enolase (NSE), and
glial fibrillar acidic protein (3–5).
NSE is a marker of acute neuronal damage (3). It rises
acutely after stroke and anoxia. Enolase is a glycolytic
enzyme found in three forms, ?, ?, and ?. The ? fraction
is found in high concentration in neuroectodermal tissue
and is known as NSE.
Most non-neuronal tissues contain a rather low level
of NSE (5–12 ng/mL). Body fluids such as serum and
cerebrospinal fluid contain very low levels of NSE. The
normal NSE levels are 5–12 ng/mL in serum and 20
ng/mL in cerebrospinal fluid (6). Studies show that cere-
brospinal fluid NSE increases in head trauma and stroke.
Therefore, it is possible that serum NSE levels may be a
clinically useful marker of neuronal damage (4,7).
The aim of this study was to classify patients accord-
ing to their initial Glasgow Coma Score (GCS) and
measure their initial plasma NSE levels. This study ex-
amined the correlation between initial GCS-defined head
trauma groups and initial NSE levels, and also examined
the correlation between initial NSE levels and final out-
comes as measured by the Glasgow Outcome Scores
(GOS) within a given head trauma group.
MATERIALS AND METHODS
This was a prospective study with 80 trauma patients
presenting to the ED of Karadeniz Technical University
between November 1, 2003 and May 1, 2004. Patients
over the age of 18 years presenting within the first 24 h
post-trauma were enrolled in the study. Patients were
divided into four groups of 20 patients each. The first
group consisted of patients with general body trauma, but
no head trauma. The second group consisted of patients
with minor head trauma (GCS ? 13). The third group of
patients had moderate head trauma (GCS 9–13) and the
fourth group of patients had severe head trauma (GCS ?
9). Enrollment in a given group stopped as soon as 20
patients had been enrolled. Enrollment in the study con-
tinued until there were 20 patients in each group.
Venous serum samples were taken from patients at
time of presentation, centrifuged, and stored at ?22°C.
Patients’ age and gender were recorded at admission and
their GCS were determined on arrival, and GOS were
determined 30 days after injury. NSE levels were mea-
sured using the presentation venous serum samples. A
Modular Analytics E170 (Roche Diagnostics, Basel,
Switzerland) immunoassay analyzer was used to deter-
mine NSE levels that were reported in ng/mL. Our lab-
oratory and international standard normal NSE range is
Surviving patients were scheduled for neurosurgery
clinic follow-up appointments 1 month later. At that time
Table 1. Clinical Characteristics of Patients in the Total
Trauma Group (n ? 80)
Mechanism of injury
Motor vehicle collision
Auto pedestrian collision
18–88 (31) years
Table 2. The Patients’ Arrival and Discharge Glasgow Coma Scale (GCS) Scores
On Admission: Number of Patients On Discharge: Number of Patients
14131211 109876543 Died15
14 13 1211109876543Died
Group 1 ? General Trauma without Head Trauma Group; Group 2 ? Mild Head Trauma Group; Group 3 ? Moderate Head Trauma
Group; Group 4 ? Severe Head Trauma Group.
298 E. Meric et al.
their GOS scores were evaluated. Patients who failed to
attend the follow-up appointment were contacted by tele-
phone, and their GOS scores were determined. Those
who evaluated the GOS were blinded to the initial GCS
and NSE levels.
We investigated the relationship between GSC-
defined head trauma groups and initial serum NSE levels.
We also investigated the relationship between initial
NSE levels and final GOS results within each head
trauma group. Finally, we examined the relationship
between NSE and GCS within each head trauma group.
Data were statistically analyzed using the non-parametric
correlation test and the post hoc test.
Eighty patients were studied prospectively. They pre-
sented to the university hospital ED with either general
trauma or head trauma over a 6-month period and met
the study criteria. The patients’ characteristics are de-
scribed in Tables 1 and 2.
NSE Results by GCS-Defined Head Trauma Group
The mean NSE level for group 1 (general trauma) was
7.54 ng/mL. For group 2 (mild head trauma), it was
19.62 ng/mL. For group 3 (moderate head trauma), it
was 54.52 ng/mL. Finally, for group 4 (severe head
trauma), it was 81.3 ng/mL. There was no significant
difference in NSE levels between groups 1 (general) and
2 (mild) (p ? 0.05), but there was a significant difference
in NSE levels between group 1 (general) and group 3
(moderate). There was also a significant difference in
NSE levels between groups 1 (general) and 4 (severe)
(p ? 0.05), between groups 2 (mild) and 3 (moderate)
(p ? 0.05), and between groups 3 (moderate) and 4
(severe) (p ? 0.49).
NSE Levels within a Head Trauma Group Compared
with GOS Results (Figures 2, 3)
When NSE levels were compared with their GOS results
within each specific initial GCS-defined head trauma
group, it was determined that GOS scores fell as NSE
levels rose. Although there was no statistically signifi-
cant difference in NSE levels with GOS within group 2
(mild head trauma) (p ? 0.05), NSE did vary signifi-
cantly with GOS within group 3 (moderate head trauma)
and within group 4 (severe head trauma) (p ? 0.05).
NSE Levels within a GCS-Defined Head Trauma
Group (Figure 4)
When patients’ arrival GCS scores and NSE levels were
compared within each head trauma group, it was deter-
mined that GCS tended to fall as NSE levels rose. But this
correlation of admit NSE and admit GCS was not statisti-
cally significant within groups 1 (general), 3 (moderate),
and 4 (severe) (p ? 0.05). There was, however, a statisti-
cally significant correlation between NSE and admit GCS
within group 2 (minor head trauma) (p ? 0.05).
Figure 1. Group’s neuron-specific enolase (NSE) levels.
Figure 2. A comparison of the moderate head trauma
group’s neuron-specific enolase (NSE) values and Glasgow
Coma Scores (GCS).
Figure 3. A comparison of the severe head trauma group’s
neuron-specific enolase (NSE) values and Glasgow Out-
come Scores (GOS).
Neuron-specific Enolase in Head Trauma Patients299
NSE Levels as a Predictor of Poor Neurologic
Outcome (Figure 5)
A ROC analysis of serum NSE and poor neurologic
outcome was done with combined data from all patient
groups. With unsatisfactory neurologic outcome defined
as GOS ? 3 and a serum NSE cutoff value of 20.52
ng/mL, serum NSE was 87% sensitive and 82.1% spe-
cific in predicting poor neurologic outcome in the study
patients. The area under the curve was 0.931.
Several experimental and clinical studies have been con-
ducted over the last 15 years to determine the practicality
of NSE as a marker of traumatic brain injury and patient
prognosis (4,8,9). However, the results of these studies
reveal no consistent correlation between NSE and the
severity of brain injury.
We prospectively studied admission NSE levels, ad-
mission and discharge GCS, and final GOS in 80 adult
In the general trauma without head trauma group,
NSE levels were not elevated above normal levels. When
compared to the general trauma group, the mild head
trauma group showed a small, but not statistically sig-
nificant increase in NSE. However, a significant eleva-
tion of NSE was found in both the moderate and severe
head trauma groups.
There was also a significant increase in serum NSE
levels in the severe head trauma group when compared
with the moderate head trauma group.
Within both the severe and moderate head trauma
group, there was a statistically significant inverse corre-
lation between GOS and NSE. There was no correlation
between GOS and NSE within the minor head trauma
We found a statistically insignificant increase in NSE
levels in the minor trauma group. There was a statisti-
cally significant inverse correlation between NSE level
and GCS within the mild head trauma group, but not
within the other groups. There was no significant rela-
tionship between NSE and GOS in the mild head trauma
group. In a controlled study of minor head trauma pa-
tients, de Kruijk et al. found insignificant elevations of
NSE. In the same patients, he found statistically signif-
icantly elevated levels of S-100B protein levels (10).
Another study, by Ross et al., showed no correlation
between serum NSE levels and GCS values or GOS in
major head trauma (11). When we examined the moder-
ate and severe head trauma groups, NSE levels were
significantly elevated compared to the no head trauma
group. A total of 18 (45%) patients from these two
groups died. This finding is compatible with those of
previous studies (1,12). Vos et al. reported high post-
trauma NSE and S-100B protein values in severe head
trauma patients. There was a statistically significant cor-
relation between NSE and S-100B levels when compared
with CT results, but not with GCS (1). In a study of 47
patients with isolated head trauma, Kuroiwa et al. found
significantly elevated serum NSE levels in patients with
moderate and severe head trauma (13).
In our study, there was a statistically significant in-
verse correlation between serum NSE levels and GOS 1
month later within the moderate and severe head trauma
groups. There was no significant association between the
NSE and GOS within the minor head trauma group.
Moreover, with unsatisfactory neurologic outcome de-
fined as GOS ? 3 and a serum NSE cutoff value of 20.52
ng/mL, serum NSE was 87% sensitive and 82.1% spe-
cific in predicting poor neurologic outcome in all study
Figure 4. A comparison of minor head trauma patients’ Glas-
gow Outcome Scores (GOS) and neuron-specific enolase
Figure 5. Neuron-specific enolase (NSE) levels as a predictor
of poor neurologic outcome.
300E. Meric et al.
patients. This derived cutoff value now needs to be Download full-text
prospectively validated. Ross et al. also found a statisti-
cally significant correlation between serum NSE levels
and GOS 1 month later (11). In contrast to our study, a
study by Raabe et al. demonstrated no significant corre-
lation between NSE levels and GOS 6 months later. The
Raabe study was different in design, in that only patients
with severe head trauma were enrolled, and NSE levels
were taken every 24 h for 10 days (14).
Hardemark et al. found that cerebrospinal fluid
S-100B protein and NSE levels were significantly higher
in an experimental contusion model when compared to
the control group (15). The first spike in NSE levels
occurred after 7.5 h in that study. They found a second
NSE spike after 1.5 days and attributed this to secondary
One of the main factors restricting the use of NSE as
a marker in brain injury is that NSE is present in eryth-
rocytes in addition to neuronal tissue. As a result, NSE
results may be misleading in patients with hemolysis.
Because NSE has a long biological half life (? 20 h),
NSE results may be affected by sample timing. We
attempted to control for this by obtaining our NSE sam-
ples within the first 24 h post trauma.
Our study did not control for mechanism of injury. Also,
our study did not control for patient age, which could
have affected neurologic outcome. We attempted to con-
trol for the effect of time from trauma on serum NSE
levels by drawing all samples within the first 24 h after
trauma. However, due to the variable time from injury to
patient presentation or transfer to our ED, we were
unable to precisely control the time from injury to NSE
level determination. This variability could have affected
In moderate and severe head trauma patients, if signifi-
cant hemolysis can be excluded, serum NSE levels taken
within the first 24 h after trauma may provide significant
additional prognostic information compared to GCS and
neuroradiologic examinations alone. A larger study with
measurements of other trauma severity scores may be
useful. It would be valuable to determine if NSE levels
are still helpful for determining neurologic prognosis in
patients with varying degrees of severe blunt trauma. The
amount of hemolysis that occurs with major blunt trauma
may negate the usefulness of NSE as a neurologic out-
come marker. Also, serum NSE was 87% sensitive and
82.1% specific in predicting poor neurologic outcome in
all of the study patients. This derived cutoff value now
needs to be prospectively validated.
1. Vos PE, Lamers K, Hendriks J, et al. Glial and neuronal proteins
in serum predict outcome after severe traumatic brain injury.
2. Hermann M, Curio N, Jost S, Wunderlich MT, Synowitz H,
Wallesch CW. Protein S-100B and neuron specific enolase as early
neurobiochemical markers of the severity of traumatic brain injury.
Restor Neurol Neurosci 1999;14:109–14.
3. Heck CN. Markers of neuronal injury in status epilepticus. Epi-
lepsy Q 1996;37:606–9.
4. Berger RP, Adelson PD, Richichi R, Kochanek PM. Serum bi-
omarkers after traumatic and hypoxemic brain injuries: insight into
the biochemical response of the pediatric brain to inflicted brain
injury. Dev Neurosci 2006;28:327–35.
5. Kukacka J, Vajtr D, Huska D, et al. Blood melallothionein, neuron
specific enolase, and protein S100B in patients with traumatic
brain injury. Neuro Endocrinol Lett 2006;27(Suppl 2):116–20.
6. Marangos PJ, Schmechel DE. Neuron specific enolase, a clinically
useful marker for neurons and neuroendocrine cells. Annu Rev
7. DeGiorgio CM, Gott PS, Rabinowicz AL, Heck CN, Smith TD,
Correale JD. Neuron-specific enolase, a marker of acute neuronal
injury, is increased in complex partial status epilepticus. Epilepsia
8. Bakay RA, Sweeney KM, Wood JH. Pathophysiology of cerebro-
spinal fluid in head injury: part 2. Biochemical markers for central
nervous system trauma. Neurosurgery 1996;18:376–82.
9. Schoerkhuber W, Kittler H, Sterz F, et al. Time course of serum
neuron-specific enolase. A predictor of neurological outcome in
patients resuscitated from cardiac arrest. Stroke 1999;30:1598–
10. de Kruijk JR, Leffers P, Menheere PP, Meerhoff S, Twijnstra A.
S-100B and neuron-specific enolase in serum of mild traumatic
brain injury patients. A comparison with health controls. Acta
Neurol Scand 2001;103:175–9.
11. Ross SA, Cunningham RT, Johnston CF, Rowlands BJ. Neuron-
specific enolase as an aid to outcome prediction in head injury. Br J
12. Yamazaki Y, Yada K, Morii S, Kitahara T, Ohwada T. Diagnostic
significance of serum neuron-specific enolase and myelin basic
protein assay in patients with acute head injury. Surg Neurol
13. Kuroiwa T, Tanabe H, Takatsuka H, Arai M, Nagasawa S, Ohta T.
Significance of serum neuron-specific enolase levels after head
injury. No Shinkei Geka 1993;21:1021–4.
14. Raabe A, Grolms C, Seifert V. Serum markers of brain damage and
outcome prediction in patients with severe head injury. Br J Neu-
15. Hardemark H, Ericsson N, Kotwica Z, et al. S-100 protein and
neuron-specific enolase in CSF after experimental traumatic or
focal ischemic brain damage. J Neurosurg 1995;71:727–31.
Neuron-specific Enolase in Head Trauma Patients301